Peter has lipoprotein(a), or Lp(a), a genetic pattern shared by 11% of Americans.
It means that Peter inherited a gene that codes for a protein, called apoprotein(a), that attaches to LDL particles, forming the combined particle Lp(a). It also means that his overall pattern responds well to a high-fat, high-protein, low-carbohydrate diet: The small LDL particles that accompany Lp(a) over 90% of the time are reduced, Lp(a) itself is modestly reduced, other abnormalities like high triglycerides (that facilitate Lp(a)'s adverse effects) are corrected. Small LDL particles are, by the way, part of the genetic "package" of Lp(a) in most carriers.
Peter also has another gene for Apo E4, another genetically-determined pattern shared by 19% of Americans. (Another 2% of Americans have two "doses" of Apo E4, i.e., they are homozygotes for E4.) This means that the Apo E protein, normally responsible for liver uptake and disposal of lipoproteins (especially VLDL), is defective. In people with Apo E4, the higher the fat intake, the more LDL particles accumulate. (The explanation for this effect is not entirely clear, but it may represent excessive defective Apo E-enriched VLDL that competes with LDL for liver uptake.) People with Apo E4 therefore drop LDL (and LDL particle number and apoprotein B) with reductions in fat intake.
This is a genetic rock-and-a-hard-place, or what I call a genetic incompatibility. If Peter increases fat and reduces carbohydrates to reduce Lp(a)/small LDL, then LDL measures like LDL particle number, apoprotein B, and LDL cholesterol will increase. Paradoxically, sometimes small LDL particles will even increase in some genetically predisposed people.
If Peter decreases fat and increases carbohydrates, LDL particle number, apoprotein B, and LDL cholesterol will decrease, but the proportion of small LDL will increase and Lp(a) may increase.
Thankfully, such "genetic incompatibilities" are uncommon. In my large practice, for instance, I have about 5 such people.
The message: If you witness paradoxic responses that don't make sense or follow the usual pattern, e.g., reductions in LDL particle number, apoprotein B, and small LDL with reductions in their dietary triggers (i.e., carbohydrates, especially wheat), then consider a competing genetic trait such as Apo E4.